Skip to content

Latest commit

 

History

History
107 lines (70 loc) · 2.92 KB

README.md

File metadata and controls

107 lines (70 loc) · 2.92 KB

Stable Diffusion Models Auto Downloader

Jupyter notebook for easily downloading Stable Diffusion models (e.g. checkpoints, VAEs, LoRAs, etc).

This notebook is developed to use services like runpod.io more conveniently.

But in fact, you can use this notebook in any environments (local machine, cloud server, Colab, etc).

Features

  • Provides Frequently Used Models (e.g. ControlNet) As Preset (Checkbox)
  • Custom Download URL
  • Automatic Download Of Preview Image
  • Automatically Converts Normal Page URL To Download URL
  • Supports Sub-Directory (custom category support)
  • Supports Any Project (e.g. stable-diffusion-webui, ComfyUI, etc)
  • Supports Auto Installation Of Extensions

Usage

Step 1

Just simply clone the project.

git clone https://github.com/jjangga0214/sd-models-downloader.git

Then open the Jupyter Notebook sd-models-downloader/index.ipynb.

Step 2

A section like the screenshot below would appear. Execute the cell.


./images/step-2.png


Step 3

Now the UI is generated. (The screenshot below is just part of the whole page.)

There're sections for Checkpoints, VAEs, Textual Inversions, Hyper Networks, LoRA, LyCORIS, ContrnolNet (v1.0, v1.1), T2I-Adapter, CoAdapter, and Extensions.


./images/step-3.png


Step 4

As the UI page (output of the cell) is long, it may be trimmed. In that case, click the message ("SHOW MORE OUTPUTS") like the screenshot below. It will be at the end of the UI. Then the full UI will be shown.


./images/step-4-1.png


The message can be different by the environment. For example, it might be like this screenshot below.


./images/step-4-2.png


Step 5

Now it's time to download. Simply execute the cell!


./images/step-5.png


Step 6

When you notice you need more models later, then just change the model selection and execute the download cell again.

Trouble Shooting

Timeout

Sometimes the download request is timed out.

./images/trouble-shooting-timeout.png

This is not a bug at all. The server may be busy or the network is congested. CivitAI has a bit higher probability than others like Huggingface. In this case, just try again! When the re-trial also fails, you should wait until the server or network gets healthier. If you don't want to wait, change the URL.

Infinite Awaiting

Sometimes download is blocked by infinite awaiting.

./images/trouble-shooting-awaiting.png

This is not a bug either. You should just wait a minute at first. But if it doesn't change, you should stop the cell and take an action. The solution is the same as the "Timeout" section above.

License

MIT License. Copyright © 2023, GIL B. Chan <github.com/jjangga0214> [email protected]